The development of Mellanox/NVIDIA GPUDirect over InfiniBand—a new model for GPU to GPU communications

Gilad Shainer1, Ali Ayoub2, Pak Lui2, Tong Liu2, Michael Kagan2, Christian Robert Trott3, Greg Scantlen4, Paul Crozier5
1HPC Advisory Council, Sunnyvale, USA
2Mellanox Technologies, Sunnyvale, USA
3Institut für Physik, Technische Universität at Ilmenau, Ilmenau, Germany
4Creative Consultants, Albuquerque, USA
5Sandia National Laboratories, Albuquerque, USA

Tóm tắt

Từ khóa


Tài liệu tham khảo

Kindratenko V, Enos J, Shi G, Showerman M, Arnold G, Stone J, Phillips J, Hwu W-m (2009) GPU clusters for high-performance computing. In: Cluster computing and workshops

Wu E, Liu Y (2008) Emerging technology about GPGPU. In: Circuits and systems

Chen G, Li G, Pei S, Wu B (2009) High performance computing via a GPU. In: Information science and engineering (ICISE)

Garland M (2010) Parallel computing with CUDA. In: Parallel & distributed processing (IPDPS)

The TOP500 list— www.top500.org

InfiniBand Trade Association— www.infinibandta.org/

Mellanox Technologies— www.mellanox.com

LAMMPS and the USER-CUDA package— http://lammps.sandia.gov/ , http://code.google.com/p/gpulammps/